Objective: The objective of this study was to develop a new 2-regression model relating Actical activity counts to METs.
Methods: Forty-eight participants ((mean±SD) age: 35±11.4 yrs) performed 10-min bouts of various activities ranging from sedentary behaviors to vigorous physical activities. Eighteen activities were split into three routines with each routine being performed by 20 individuals. Forty-five routines were randomly selected for the development of a new 2-regression model and 15 tests were used to cross-validate the new 2-regression model and compare it against existing equations. During each routine, the participant wore an Actical accelerometer on the hip and oxygen consumption was simultaneously measured by a portable metabolic system. The coefficient of variation (CV) of four consecutive 15-sec epochs was calculated for each minute. For each activity, the average CV and the counts.min-1 were calculated for minutes 4-9. If the CV was ≤ 13% a walk/run regression equation was used, and if the CV was > 13% a lifestyle/leisure time physical activity regression was used.
Results: An exponential regression line (R2=0.912; SEE=0.149) was used for activities with a CV ≤ 13%, and a cubic regression line (R2=0.884, SEE=0.804) was used for activities with a CV > 13%. In the cross-validation group the mean estimates, using the new 2-regression model with an inactivity threshold, were within 0.56 METs of measured METs for each of the activities performed (P≥0.05), except cycling (P<0.05).
Conclusion: For most activities examined the new 2-regression model predicted METs more accurate than currently available equations for the Actical accelerometer.
- Motion Sensor
- activity count variability
- oxygen consumption
- physical activity